Add Harper workflow for grammar and style checks

.github/workflows/harper.yml

@@ -0,0 +1,114 @@
+name: Harper
+
+on:
+  pull_request:
+  workflow_dispatch:
+
+jobs:
+  harper:
+    name: Grammar and Style Check
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v6
+
+      - name: Cache Cargo
+        uses: actions/cache@v4
+        with:
+          path: |
+            ~/.cargo/bin
+            ~/.cargo/registry
+            ~/.cargo/git
+          key: ${{ runner.os }}-cargo-harper
+          restore-keys: |
+            ${{ runner.os }}-cargo-harper
+
+      - name: Install Harper CLI
+        run: |
+          if ! command -v harper-cli >/dev/null 2>&1; then
+            cargo install --git https://github.com/Automattic/harper harper-cli --locked
+          else
+            echo "Harper CLI found in cache."
+          fi
+
+      - name: Generate Harper JSON Report
+        run: |
+          find docs src/pages -type f \( -name "*.md" -o -name "*.mdx" \) -print0 | \
+          xargs -0 harper-cli lint \
+            --user-dict-path ./.harper-dictionary.txt \
+            --format json \
+            > harper-report.json || true
+
+      - name: Extract Spelling Candidates
+        run: |
+          python - <<'PY'
+          import json
+
+          with open("harper-report.json", encoding="utf-8") as f:
+              reports = json.load(f)
+
+          words = set()
+
+          for report in reports:
+              for lint in report.get("lints", []):
+                  kind = lint.get("kind")
+                  rule = lint.get("rule")
+                  word = lint.get("matched_text")
+
+                  if kind == "Spelling" and rule == "SpellCheck" and word:
+                      words.add(word)
+
+          print("Suggested dictionary candidates:")
+          for word in sorted(words, key=str.lower):
+              print(word)
+          PY
+
+      - name: Check Selected Harper Rules
+        run: |
+          python - <<'PY'
+          import json
+          import sys
+
+          with open("harper-report.json", encoding="utf-8") as f:
+              reports = json.load(f)
+
+          selected = []
+          total_lints = 0
+
+          IGNORED_RULES = {
+              ("Spelling", "OkToOkay"),
+              ("Spelling", "DisjointPrefixes"),
+          }
+
+          for report in reports:
+              file = report.get("file")
+
+              for lint in report.get("lints", []):
+                  total_lints += 1
+
+                  kind = lint.get("kind")
+                  rule = lint.get("rule")
+                  line_no = lint.get("line")
+                  message = lint.get("message")
+                  text = lint.get("matched_text")
+
+                  if (kind, rule) in IGNORED_RULES:
+                      continue
+
+                  if kind in {"Spelling", "Grammar", "Repetition"}:
+                      selected.append(
+                          f"{file}:{line_no}: [{kind}::{rule}] {text!r} - {message}"
+                      )
+
+          print(f"Total Harper lint count: {total_lints}")
+          print(f"Selected lint count: {len(selected)}")
+
+          if selected:
+              print("Selected Harper lints found:")
+              for item in selected:
+                  print(item)
+              sys.exit(1)
+
+          print("No selected Harper lints found.")
+          PY